Wire Arc Additive Manufacturing (WAAM) has been changing high-end manufacturing by building metal components layer by layer. This process, which allows for the production of large and complex parts, has attracted significant industry attention due to its ability to produce metal components with reduced lead times and minimal material waste. During WAAM part fabrication, ensuring accurate bead geometry is essential for minimizing defects and improving overall part quality. Achieving this requires a closed-loop control system to enhance accuracy in bead geometry throughout the manufacturing process. The design of such a control system requires a bead geometry model under varying process parameters. However, modeling this is challenging due to the complex and nonlinear interactions among these parameters. To address this challenge, this paper presents a LSTM-based (Long Short Term Memory) approach for estimating the bead width in WAAM. The method involves creating a dataset by varying WAAM process parameters, such as travel speed and wire feed speed, on a welding system that includes both a welding power source and a robotic manipulator. A high-dynamic-range camera captures real-time images of the weld bead, and an image-processing algorithm estimates the bead width, synchronizing this data with signals from the welding system. The model hyperparameters are optimized through Bayesian optimization applied to the cross-validation process. The selected model is integrated into a closed-loop control system with an integral controller to regulate bead width. Initially, the model is employed for tuning and testing the controller in a simulated environment. The tuned controller is then applied in a real experiment, where the results demonstrates strong similarity to the desired outcomes.

This paper presents an accurate and computationally efficient stairway detection algorithm, robust to measurement noise. While most similar methods depend heavily on detecting three-dimensional (3D) features, this work focuses on locating two-dimensional (2D) features embedded in 3D data. The main idea is: each stair is a sequence of lines with similar domain of definition, and the staircase is a sequence of such elements. To avoid unnecessary computations, an octree is used to down-sample the environment point-cloud data. The resulting centroids are stored in n arrays, each containing points with equal height. This effectively segments the 3D data into n 2D slices. The Hough transform is used to scan each array for lines. the remainder of the algorithm detects lines with the same domain of definition and determines if sequences that may correspond to stairways exist. If these sequences meet the required tolerances, then a staircase is detected. For each detected stairway a model that contains the width, height, depth, and amount of stairs is returned. The proposed algorithm is able to efficiently locate and model staircases on different poses and is robust to noise from the measured data, as indicated by the experimental results.

This paper presents a novel control architecture for controlling a moving remote center of motion in addition to the end-effector motion during robotic surgery. In minimally invasive surgery, it is common to require that the point at which the robot enters the body, called the incision point or the trocar, does not allow for any lateral motion. It is generally considered that no motion should be applied to this point in order to avoid inflicting damage to the patient's skin. However, in surgery, the patient's body may be moving, for example due to breathing or the beating of the heart. In order to compensate for this motion-or if we for some other reason want to leverage the possible motion of the incision point to improve performance in any other way-we derive a new framework which allows us to actively control the motion both at the incision point and the end effector. The novelty of the approach lies in the possibility of controlling both the incision point and the end effector to follow a trajectory, and that we find a Jacobian matrix that satisfies the velocity constraints in both the end-effector and the incision point frames. This allows us to formulate a framework that is not only suited for control, but also for analyzing the condition number of the Jacobian and avoid any singular configurations that may arise either as a result of the constrained motion or the manipulator geometry. The approach is verified experimentally on a redundant industrial manipulator.

Mobile robots and vehicles with active articulated elements are well suited to drive on irregular and rough terrain because they are able to adjust their center of mass and decrease the risk of tip over or other accidents. The articulated mechanism, however, is usually constrained by the actuator's bandwidth, making it difficult for the vehicle to compensate for abrupt changes in terrain profile. In this paper we propose a terrain model-based control method to improve stability when traversing terrains with varying slopes, depressions, and rises. A model predictive control approach takes into account actuator bandwidth to anticipate interactions with the terrain and control the articulated elements to prevent tipping over. The method is applied to an autonomous orchard platform where workers stand while conducting production tasks on the trees, adjusting the platform height and increasing its stability. The feasibility is illustrated via numerical simulations performed with the MD Adams/Car software and Matlab, with both artificial and natural terrain data. Challenging yet realistic terrain profiles are considered to demonstrate the control effectiveness in preventing the platform from tipping over and therefore improving the workers' and the vehicle's safety.

This paper considers the active control problem of reconfigurable mobile robots on irregular terrain. A kinematic control strategy to improve robot mobility (stability and traction) is proposed. The proposed control is validated through numerical simulations and experimental tests using an amphibious wheel-legged robot, named Environmental Hybrid Robot, recently developed by Petrobras S.A. (Brazilian Petroleum Company) for environmental monitoring in the Amazon rain forest.

This work considers a hybrid force and vision control system for robotic manipulators using a force sensor and a fixed uncalibrated camera. A method is proposed to combine direct force control and adaptive visual servoing to perform tasks on unknown smooth surfaces, in the presence of uncertainties in the camera-robot system parameters. The considered task involves the visual tracking of a moving target, while the end-effector tip exerts a controlled contact force on the surface. Simulation results are presented to illustrate the performance of the proposed scheme

In this work, a new 2D adaptive visual positioning/tracking controller is considered. The proposed scheme is developed for image-based eye-to-hand systems to perform position and tracking tasks on the 3D environment, when both camera calibration and robot parameters are uncertain. The symmetric-diagonal-upper (SDU) factorization method is adopted to solve the adaptive multivariable control problem. The positioning/tracking controller is first developed for the Cartesian robot case, and then extended to the general robotic system. Simulation results are also presented for the proposed strategy.

A control-theoretical solution is presented for the direct adaptive visual control of planar manipulators using a fixed camera, when both camera calibration and robot dynamics are uncertain. The proposed scheme is developed for image-based look-and-move visual systems to allow tracking of a moving target. In order to solve the multivariable parameter adaptive problem, the proposed immersion and invariance (I&I) method is used. The scheme is then combined with the adaptive controller for the manipulator, taking into account its nonlinear dynamics and leading to an overall stable adaptive visual system. Simulations and experimental results are also presented for the proposed strategy.

This paper considers the localization and navigation of a mobile robot. The control strategy is based on a nonlinear model predictive control technique that utilizes the Newton method. The robot localization is obtained using information from odometric and ultrasonic sensors through a Kalman filter. Simulation and experimental results illustrate the efficacy of the proposed method.

Describes the development and implementation of a dynamic positioning system for remotely operated underwater vehicles (ROVs) which is based on a mechanical passive arm (PA) for position measurement. The system, named DPSROV, is suited for tasks which require precise positioning within a limited range around a working site. The DPSROV was configured aiming at easy installation in most commercially available ROVs. The effectiveness of the DPSROV was tested on real ROVs. Two different control strategies were adopted: a conventional linear controller and a robust variable structure model-reference adaptive control. Experimental results are presented to illustrate the DPSROV performance.

A class of iterative methods have recently been proposed for the path planning of nonholonomic systems. These methods warp an initial path iteratively to an acceptable final path by using Newton-Raphson or gradient type of algorithms. Once a path is found off-line, a feedback controller is then used to follow the path. In this paper, we propose a modification of these off-line methods to transform them directly into a feedback controller. The main idea is to couple the iteration variable to the actual time, thus the control is executed during the path iteration, before the convergence. We show that this scheme guarantees the closed loop asymptotic stability when the system model is known, and possesses certain robustness when the model information is imperfect. By using interior penalty functions, inequality constraints can also be handled by the algorithm. Simulation results are included, showing promise of the approach.

It is well known that the linear feedback of the quarternion of the attitude error and angular velocity globally stabilizes the attitude of a rigid body. In this note, we show that the angular velocity feedback can be replaced by a nonlinear filter of the quarternion, thus removing the need of direct angular velocity measurement. In contrast to other approaches, this design exploits the inherent passivity of the system; a dynamic observer reconstructing the velocity is not needed. An application of the proposed scheme is illustrated for the robot control problem. Simulation results are included to illustrate the theoretical results.